Introduction

In recent years noninvasive positive pressure ventilation (NPPV) has gained increasingly widespread acceptance for the support of chronic and acute ventilatory failure [1, 2]. Continuous positive airway pressure (CPAP) is the simplest form of NPPV [1, 2], and its efficiency has been confirmed in common clinical conditions such as cardiogenic pulmonary edema [3], chronic obstructive pulmonary disease [4], and chest wall trauma [5]. CPAP can be provided with continuous flow generators [1, 6], high-pressure driven ventilators [1, 6], turbine-driven CPAP machines [7], and most of ventilators designed to administer noninvasive positive pressure ventilation (NPPV ventilator) [1]. The advantages of flow generators over NPPV ventilators are the fact that they are easier to set up and use, less expensive, and more portable. Unfortunately, there are no studies comparing the performance of continuous flow generators with NPPV ventilators to deliver CPAP. Our objective was to compare the performances of two different continuous flow generators supplied with different pressures with a ventilator designed to deliver NPPV.

Materials and methods

We tested two flow generators the Adjustable Downs Flow Generator (Vital Signs, USA; CPAP1) [8] and the Whisperflow Variable Flow (Caradyne, Ireland; CPAP2) [9]. Both were used with a double-limb respiratory circuit with one-way exhalation isobaric spring-load CPAP valves without flow dependence [10] (Vital Signs, USA). We also tested the BiPAP Vision (NPPV ventilator; Respironics, USA) with its disposable exhalation port incorporated into circuit.

We employed a mechanical lung model modified from previous studies [11, 12]. Briefly, the experimental set used a two-chambered training test lung (Adult TTL, Michigan Instruments, USA) with one chamber connected to a trigger ventilator (Bear 1000,Viasys, USA) and the other to a mannequin head (C500, Kapta, Brazil). A facial mask fitted on mannequin head was connected to the flow generators or NPPV ventilator. The compliance of the lung simulator was 50 ml/cmH2O and airway resistance 5 cm H2O l−1 s−1. An airway pressure transducer (DP45-30, Valydine, USA) and a pneumotachograph (Flow head 3700, Hans Rudolph, USA) were connected to the tubes inside the mannequin. Signals were sampled at 200 Hz for off-line analysis.

The performance of the NPPV ventilator and flow generators was compared at three CPAP levels (5, 10, and 15 cmH2O). Flow generators were used at 100, 200, and 300 kPa through a reducing-pressure valve (Moriya 700810, Brazil) connected to hospital oxygen supply system. Flow generators were used with 100% of inspiratory oxygen fraction and the flow adjustment valve fully on. The NPPV ventilator did not deliver pressure support during CPAP (Fig. 1). At each CPAP level we analyzed two different inspiratory efforts. Lung simulator was set at tidal volume of 400 ml and inspiratory flow of 24 l/min to simulate a low inspiratory effort (LE) and 800 ml with 48 l/min to high effort (HE). At each CPAP level and inspiratory effort we recorded ten respiratory cycles, and performance was compared by differences in three clinical relevant variables: (a) the area under the adjusted CPAP (areaCPAP), defined as the pressure area under the CPAP level from the onset to the end of inspiratory flow (ideally this should be zero, Fig. 1, because CPAP level must be maintained constant to provide optimal assistance) [13], (b) the capacity to attain the preset CPAP, defined as the competence of the flow generator or NPPV ventilator to achieve the preset pressure [14], and (c) tidal volume, calculated as the time integration of inspiratory flow. Flow generators and the NPPV ventilator output flows were measured with a calibration analyzer (RespiCal-Timeter, Allied Health Care, USA).

Fig. 1
figure 1

Pressure and flow curves. This sample was obtained during the recording of flow and continuous positive airway pressure (CPAP) level at CPAP of 15 cmH2O in the high inspiratory effort. The increase in the CPAP level at the beginning of the expiration is a noise of our model and is caused by an asynchrony of test lung chambers during inspiration to expiration transition

Full size image

Data are shown as the mean of ten consecutive respiratory cycles. The measures were very stable along the recording time and the variances of most of measurements were null or negligible (see Electronic Supplementary Material), and therefore no statistical approach but straight comparison was used for analysis.

Results

CPAP1 and CPAP2 output flows increased with the increment in the pressure supply but did not change with different inspiratory efforts. NPPV ventilator output flow varied with the CPAP level and inspiratory effort (Table 1). With low inspiratory effort, independently of the CPAP level, the areaCPAP values between devices were close, except for CPAP2 at 100 kPa which had a poorer performance (Fig. 2a). In the high inspiratory effort CPAP1 performance was better because its areaCPAP was always smaller than that of CPAP2 and NPPV. Independently of the level of inspiratory effort, with the increase in the pressure supply, CPAP1 increased areaCPAP while CPAP2 decreased it (Fig. 2). NPPV ventilator and CPAP1 were able to attain the preset pressure of 5, 10, and 15 cmH2O. CPAP1 overshot the preset CPAP at 200 and 300 kPa. CPAP2 was able to attain the CPAP of 5 cmH2O but was unable to attain 10 cmH2O at 100 kPa ornd 15 cmH2O at any pressure supply (Fig. 3). With low and high inspiratory effort, independent of CPAP level and pressure supply, all devices generated similar tidal volumes. For both flow generators the best results regarding areaCPAP and capacity to attain the preset CPAP were achieved with output flow around 100 l/min (123 l/min at 100 kPa for CPAP1 and 94 l/min at 300 kPa for CPAP2; Fig. 4).

Table 1 Flow generators and NPPV ventilator output flows
Full size table
Fig. 2
figure 2

AreaCPAP (cmH2O/s) during inspiration in low and high inspiratory effort. CPAP Continuous positive airway pressure; Area CPAp area under the adjusted CPAP; CPAP1 Adjustable Downs Flow Generator; CPAP2 WhisperFlow Variable Flow; NPPV ventilator BiPAP Vision. a Low inspiratory effort. b High inspiratory effort

Full size image
Fig. 3
figure 3

Capacity to attain the preset continuous positive airway pressure (CPAP). CPAP1 Adjustable Downs Flow Generator; CPAP2 WhisperFlow Variable Flow; NPPV ventilator BiPAP Vision

Full size image
Fig. 4
figure 4

AreaCPAP × flow generators output flow. CPAP Continuous positive airway pressure; Area CPAp area under the adjusted CPAP; low effort low inspiratory effort; high effort high inspiratory effort. At each line, from the left to the right, the first three symbols belong to flow generator 2 and the last three to flow generator 1

Full size image

Discussion

In our lung model study performance of CPAP1 was better when adjusted to an oxygen supply of 100 kPa, while that of CPAP2 was better when adjusted to an oxygen supply of 300 kPa. Both had better performance with an output flow around 100 l/min. Adjusted to their better oxygen supply pressure, flow generators had a similar or better capacity to maintain the CPAP level, but the NPPV ventilator was more reliable to attain the preset CPAP. Tidal volumes did not differentiate the devices, but areaCPAP and capacity to attain the preset CPAP did.

Tidal volumes were equal because all devices provided output flow in excess of volume demand. Differences in output flow among devices explained the differences in areaCPAP because areaCPAP is due mainly to sufficient or insufficient gas delivery related to the need of flow during inspiration [6]. However, the use of a high compliance reservoir bag may have improved the stability of the pressure applied to the lung model [15, 16]. Unexpectedly, CPAP1 increased its areaCPAP with the increase in pressure supply. A possible reason was a notable increase in the mask air leak with the higher flow, causing difficulty to maintain the CPAP level. In addition, in our experience, the supply of flow generators with pressures higher than 300 kPa causes patients’ discomfort, high noise and a blow out of a soft rubber valve in CPAP1, a fact that has been reported previously [8].

Differences in capacity to attain the preset CPAP may be also explained by the differences in output flow. The lowest CPAP2 output flow avoided it to reach the preset CPAP, especially at 100 kPa with CPAP of 10 and 15 cmH2O. Conversely, at the highest pressure supply CPAP1 extremely high flow with the highest pressure supply overshot the preset CPAP because the valve was unable to relieve the excessive flow. Better NPPV ventilator performance to attain the preset CPAP was expected due to its feedback microprocessed system that adjusted flow according to the preset CPAP.

For both flow generators the best results regarding areaCPAP and capacity to attain the preset CPAP were achieved with output flow around 100 l/min. This finding was corroborated by a recent study [17] which compared CPAP1 delivering a high (100 l/min) and a low (50 l/min) output flow. Independently of the inspiratory effort the areaCPAP was smaller with output flow of 100 l/min.

The major limitation of the present study is that it was performed in a lung model. Another limitation is the fact that different flow generators regulations of pressure supply, oxygen inspiratory fraction, and flow adjustment valve may have changed their output flow and performance. Our measurements were very stable along the recording time and the variances of most of measurements were null or negligible. This finding was expected and desirable in a bench fully controlled study. If we had applied any statistical test, all comparisons would have been significant.

Flow generators may be the only manner to provide CPAP in low income areas. In these areas the availability of medical equipment is scarce and proven medical treatment, such as CPAP, may be not carried out. If flow generators prove efficient, they would be an alternative in these areas. As CPAP1 may work well at 100 kPa, its use could be useful in areas not provided with high-pressure oxygen supply. It is noteworthy that simple turbine-driven CPAP machines have similar acquisition cost to flow generators. However, respect to single circuit with intentional leak turbine driven CPAP machines or noninvasive ventilators, the use of a flow generator with a double limb circuit together to a high bias flow could prevent CO2 rebreathing when CPAP is set below 8 cmH2O [18].